Perl 6 - the future is here, just unevenly distributed

IRC log for #perl11, 2013-03-05

| Channels | #perl11 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
01:49 Reini joined #perl11
02:11 logie joined #perl11
02:53 travis-ci joined #perl11
02:53 travis-ci [travis-ci] perl11/potion#135 (master - 25a3e89 : Reini Urban): The build was broken.
02:53 travis-ci [travis-ci] Change view : https://github.com/perl11/potion/compare/cac51e4dadd6...25a3e891275f
02:53 travis-ci [travis-ci] Build details : http://travis-ci.org/perl11/potion/builds/5241857
02:53 travis-ci left #perl11
04:43 dalek potion: 87acf86 | rurban++ | Makefile:
04:43 dalek potion: make: fix ifdef, APPLE is now defined to 0
04:43 dalek potion: review: https://github.com/perl11/potion/commit/87acf8604e
04:45 dalek potion/p2: 25a3e89 | rurban++ | config.mak:
04:45 dalek potion/p2: Reenable cc optims, even -O3
04:45 dalek potion/p2:
04:45 dalek potion/p2: See issue #6. Apparently the old jit problem was fixed with
04:45 dalek potion/p2: 0275bfa675e7
04:45 dalek potion/p2: review: https://github.com/perl11/potion/commit/25a3e89127
04:45 dalek potion/p2: 98ebe10 | rurban++ | config.mak:
04:45 dalek potion/p2: Merge remote-tracking branch 'origin/master' into p2
04:45 dalek potion/p2: review: https://github.com/perl11/potion/commit/98ebe10357
04:45 dalek potion/p2: 87acf86 | rurban++ | Makefile:
04:45 dalek potion/p2: make: fix ifdef, APPLE is now defined to 0
04:45 dalek potion/p2: review: https://github.com/perl11/potion/commit/87acf8604e
04:45 dalek potion/p2: 90fbf22 | rurban++ | Makefile:
04:45 dalek potion/p2: Merge remote-tracking branch 'origin/master' into p2
04:45 dalek potion/p2:
04:45 dalek potion/p2: Conflicts:
04:45 dalek potion/p2: Makefile
04:45 dalek potion/p2: review: https://github.com/perl11/potion/commit/90fbf22951
04:48 broquaint joined #perl11
04:50 logie joined #perl11
05:45 travis-ci joined #perl11
05:45 travis-ci [travis-ci] perl11/potion#136 (master - 87acf86 : Reini Urban): The build is still failing.
05:45 travis-ci [travis-ci] Change view : https://github.com/perl11/potion/compare/25a3e891275f...87acf8604e4c
05:45 travis-ci [travis-ci] Build details : http://travis-ci.org/perl11/potion/builds/5246004
05:45 travis-ci left #perl11
05:47 travis-ci joined #perl11
05:47 travis-ci [travis-ci] perl11/potion#137 (p2 - 90fbf22 : Reini Urban): The build was broken.
05:47 travis-ci [travis-ci] Change view : https://github.com/perl11/potion/compare/d37aa7543935...90fbf2295175
05:47 travis-ci [travis-ci] Build details : http://travis-ci.org/perl11/potion/builds/5246022
05:47 travis-ci left #perl11
06:16 Reini joined #perl11
06:53 Reini joined #perl11
06:53 logie joined #perl11
07:15 tokuhirom joined #perl11
08:27 Reini joined #perl11
11:29 mberends joined #perl11
13:09 Reini joined #perl11
13:27 logie joined #perl11
14:13 logie joined #perl11
16:12 Reini joined #perl11
17:20 bluescreen joined #perl11
17:35 fglock_ joined #perl11
17:35 fglock_ Reini: ping
17:43 fglock_ I'm playing with redefining v8 native types to use as a possible perl5 backend - things like making the string "0" mean false
17:45 fglock_ I wonder how is your experience with p2 going - for example, do you see it doable to implement refcounting?
18:59 Reini joined #perl11
19:19 Reini joined #perl11
19:47 rurban fglock_ I do not want to implement refcounting
19:48 rurban I beleieve it's causing a lot of security and performance problems, and it's only needed to orderly provide local destruction
19:49 rurban local destruction for filehandles - well. explicitly calling close is always better
19:51 rurban but when an object goes out of scope I'll try to call all DESTROY methods instead. I have not implement that scope cleanup code yet. But it's pretty easy, easier than supporting refcounts
19:52 rurban All local objects are handled in local registers, so I just need to walk them. All lexical objects defined in upper scopes are defined in a different frame pointing forward to the upper scope.
20:04 fglock_ for example, what if the object was assigned to a parameter - in this case, the register and the parameter point to the same object
20:06 fglock_ so you can't call DESTROY on the register
20:17 bluescreen_ joined #perl11
20:25 Reini joined #perl11
20:40 Reini joined #perl11
21:58 bluescreen__ joined #perl11
22:11 bluescreen_ joined #perl11
22:16 rurban params are known thanksfully. and refs can be resolved at run-time
22:17 rurban signature handling has even it's own op in the vm
22:18 rurban I'm thinking of adding a new endscope op to the vm, to do just that. because then the compiler can add it only when needed.
22:19 Reini joined #perl11

| Channels | #perl11 index | Today | | Search | Google Search | Plain-Text | summary